OPINION - THE END IS NOT NEAR

TCS Daily, 11 April 2006
http://www.tcsdaily.com/article.aspx?id=041106A

By Tim Worstall

There's good news, more good news and then, unfortunately, some bad news, on the subject of climate change. What would you like first? Right, the good news it is then.

In all of the arguments about climate change the two questions that have always loomed largest for me are: how much of it is there likely to be? and what are we going to do about it? If it all ends up being 0.1 degrees Celsius in a century then obviously we don't do much about it and if it's going to be 10 degrees Celsius next week then we'd better get a move on.

The Kyoto Protocol was never going to be one of the things I thought we should do as it does not very much at great expense. I'm also on record here as stating that I think technology will save us, for my day job involves some contact with certain parts of the alternative energy research world and things are moving a great deal faster than the wider world seems to recognize.

Having said that (revealing my prejudices as it were) the question of how much change we're likely to see is obviously the most important. We have a number of different estimates, using different methods, and some of them push some very scary numbers indeed. I don't mean just the usual alarmists (those who say we should all be dead already from the pesticides in our baby milk, we've already drowned from the ice caps melting and so on) but even some of the more sober scientific studies say that they can't rule out 6-degree C rises, higher even. Which is why this paper is so cheering. It looks like we can rule out runaway warming purely as a result of CO2 emissions. For an easier to understand explanation try this at the blog of one of the authors.

We have a number of different ways of trying to work out the "climate sensitivity," that is, what sort of temperature change would we expect to see from a doubling of the CO2 concentration in the atmosphere? The International Panel on Climate Change (the UN's offshoot looking into all of this on our behalf) has in the past given a range of 1.5-4.5 degrees C. Various other methods have also been used and these are the ones that don't rule out those very large changes that the alarmists tell us about in the newspapers all the time. Which leads to the interesting thing noted in the new paper:

We made the rather elementary observation that these above estimates are based
on essentially independent observational evidence, and therefore can (indeed must)
be combined by Bayes' Theorem to generate an overall estimate of climate sensitivity.

So instead of wondering which of our estimates might be correct we look at all of them and come to the correct answer. This pretty much rules out the extreme outcomes and gives us, as they say, a climate sensitivity of 3 degrees C. There's still a range there but the researchers are quite clear about the fact that they didn't think that the scientific community is ready for such a low number to be announced. All of which is of course extremely good news. Even if everything else said about climate change is true, if every Friends of the Earth pamphlet is spot on in every detail, we're still not going to have runaway global warming as a result of CO2 emissions.

Excellent, the second piece of good news also shows that estimates of how much of a rise in CO2 emissions we are going to see are also too high. Ian Castles and David Henderson made the point (explained here at TCS in 2004) that there was something decidedly odd about the Special Report on Emissions Scenarios (SRES). This is the series of economic models that tries to look at how the world is going to develop over the next century and then give the tonnages of CO2, methane and so on that will be pumped out into the atmosphere. There were several substantial criticisms (the way the use of regional growth figures would have made North Korea richer than the US in 2100 was a particular delight) but perhaps the most important was the one about the use of exchange rates.

It's well known that if you use market exchange rates to compare relative levels of wealth between rich and poor countries that you'll end up overstating the differences. Things made locally and consumed locally (so called non-traded goods) will be cheaper in the poor countries for, it being a poor place, people get paid less, amongst other factors. So when we try to make such international comparisons we are supposed to use Purchasing Power Parity (PPP) exchange rates (which take account of these differences in prices) so that we measure the true gap in wealth correctly. This shouldn't have made much difference to the SRES except for the fact that most of the models assumed "convergence". That is, that most of the poor countries would end up becoming not just less poor in absolute terms but also less poor in relative terms. Well, if you measure that poverty in the first place using market exchange rates (which the SRES did) then obviously you will overstate the amount of growth that will happen to get to that convergence. That's part of the Castles/Henderson case, that the SRES assumes too much growth in the economy over the next century. This, of course, means that they're overstating the increase in emissions that the scientists then plug into their climate models.

Many were not all that taken with this argument, amongst them the Australian economist John Quiggin, and he's continued to work away at the problem, including making submissions to The Stern Review (the UK Government's look at the economics of climate change). In the course of this he's received a paper (not peer reviewed, this is a working paper) from a colleague, a W. Erwin Diewert, which tells us that there is indeed substance to the Castles/Henderson critique. Not quite as much as was originally claimed (but then they've already dialed back from their very large first claims) but large enough for this to be the conclusion:

What conclusions can we draw from the above algebra? It seems possible to draw
the following tentative conclusions:

Castles and Henderson are right to criticize the first part of the SRES modeling
strategy, which relies on market exchange rates to calculate per capita real income
differences between countries. It would be much better to use ICP PPP's for this first
part of the SRES modeling strategy. The differences between PPP's and market exchange rates can be very large so their criticism is not a negligible one.

Quiggin is right to implicitly criticize the entire SRES modeling strategy. It would
be simpler to abandon the two stage modeling strategy and make direct comparisons of energy intensities across countries and assume energy convergence rather than real income convergence.

Either way, the SRES model should be reestimated.

Now I'll have to take what these four gentlemen, Castles, Henderson, Diewert and Quiggin tell me is their conclusion slightly on trust. But they do all agree, that the end result of their collective two-year ponder over this question is that the SRES is using the wrong numbers and or methods and that the calculations need to be done again. There are differences about how much they think things will change if these sums are done again but they are (like the good academics they are) telling the IPCC that it needs to do its homework over.

But don't you think that's two pieces of good news? That climate sensitivity is less than previously thought and also that the models everyone's been using for the past five years over-estimate (to a still argued over degree) the likely emissions over the next century?

Want the bad news? The IPCC isn't going to take any notice:

In 2001 the Intergovernmental Panel on Climate Change (IPCC) published a set of
scenarios in the Special Report on Emission Scenarios (SRES). These scenarios have
been developed in a four year process with many scientists involved in the writing and the review process. The SRES scenarios played an important role in the Third Assessment Report (TAR) of the IPCC and will be used in the upcoming Fourth Assessment Report (AR4).

The 21st IPCC plenary session (November 2003) decided that no new baseline scenario would be prepared for the AR4, in view of the time it takes before new scenarios are taken up by the research community and used in publications.

AR4 is to be published in 2007. AR5, the fifth assessment report is presumably due in 2013 or thereabouts and that's the first time that the SRES models will be looked at again. Now I don't know about you but I don't think that's all that acceptable. We are (depending upon which side of the argument you are on) either facing the greatest threat to the health of the planet or we're about to spend trillions upon trillions of dollars on fixing something that doesn't actually need fixing.

Don't you think having a few guys cranking through some spreadsheets to find out which might be a good idea? Soonish?

The author is a TCS contributing writer living in Europe.